CKD Progression From the Time of Estimated GFR-Based Waitlist Eligibility and Racial Disparities in Transplant Access

Alternative creatinine-based GFR estimates in United States populations-similar performance, same gaps-is it time to move on?

This study evaluated performance of the European Kidney Function Consortium (EKFC) equation in a US cohort, comparing population-specific (EKFCPS) with race-free (EKFCRF) Q values (median normal creatinine). Both EKFCPS and EKFCRF equations showed less bias than the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) 2021 equation. The percentage of estimated glomerular filtration rate (GFR) within 30% of measured GFR was similar for CKD-EPI 2021 (79.2% [range, 78.5%-79.9%]) and EKFCRF (80.1% [range, 79.4%-80.7%]) equations but improved with the EKFCPS equation (81.1% [range, 80.5%-81.8%]), confirming utility of the EKFC equation in US populations.

Utility of the Kidney Failure Risk Equation and Estimated GFR for Estimating Time to Kidney Failure in Advanced CKD

RATIONALE & OBJECTIVE

The Kidney Failure Risk Equation (KFRE) predicts the 2-year risk of kidney failure for patients with chronic kidney disease (CKD). Translating KFRE-predicted risk or estimated glomerular filtration rate (eGFR) into time to kidney failure could inform decision making for patients approaching kidney failure.

STUDY DESIGN

Retrospective cohort.

SETTING & PARTICIPANTS

CKD Outcomes and Practice Patterns Study (CKDOPPS) cohort of patients with an eGFR<60mL/min/1.73m2 from 34 US nephrology practices (2013-2021).

EXPOSURE

2-year KFRE risk or eGFR.

OUTCOME

Kidney failure defined as initiation of dialysis or kidney transplantation.

ANALYTICAL APPROACH

Accelerated failure time (Weibull) models used to estimate the median, 25th, and 75th percentile times to kidney failure starting from KFRE values of 20%, 40%, and 50%, and from eGFR values of 20, 15, and 10mL/min/1.73m2. We examined variability in time to kidney failure by age, sex, race, diabetes status, albuminuria, and blood pressure.

RESULTS

Overall, 1,641 participants were included (mean age 69±13 years; median eGFR of 28mL/min/1.73m2 [IQR 20-37mL/min/1.73 m2]). Over a median follow-up period of 19 months (IQR, 12-30 months), 268 participants developed kidney failure, and 180 died before reaching kidney failure. The median estimated time to kidney failure was widely variable across patient characteristics from an eGFR of 20mL/min/1.73m2 and was shorter for younger age, male sex, Black (versus non-Black), diabetes (vs no diabetes), higher albuminuria, and higher blood pressure. Estimated times to kidney failure were comparably less variable across these characteristics for KFRE thresholds and eGFR of 15 or 10mL/min/1.73m2.

LIMITATIONS

Inability to account for competing risks when estimating time to kidney failure.

CONCLUSIONS

Among those with eGFR<15mL/min/1.73m2 or KFRE risk>40%), both KFRE risk and eGFR showed similar relationships with time to kidney failure. Our results demonstrate that estimating time to kidney failure in advanced CKD can inform clinical decisions and patient counseling on prognosis, regardless of whether estimates are based on eGFR or the KFRE.

PLAIN-LANGUAGE SUMMARY

Clinicians often talk to patients with advanced chronic kidney disease about the level of kidney function expressed as the estimated glomerular filtration rate (eGFR) and about the risk of developing kidney failure, which can be estimated using the Kidney Failure Risk Equation (KFRE). In a cohort of patients with advanced chronic kidney disease, we examined how eGFR and KFRE risk predictions corresponded to the time patients had until reaching kidney failure. Among those with eGFR<15mL/min/1.73m2 or KFRE risk > 40%), both KFRE risk and eGFR showed similar relationships with time to kidney failure. Estimating time to kidney failure in advanced CKD using either eGFR or KFRE can inform clinical decisions and patient counseling on prognosis.

Use of Race in Kidney Function Estimation: Lessons Learned and the Path Toward Health Justice

In 2020, the nephrology community formally interrogated long-standing race-based clinical algorithms used in the field, including the kidney function estimation equations. A comprehensive understanding of the history of kidney function estimation and racial essentialism is necessary to understand underpinnings of the incorporation of a Black race coefficient into prior equations. We provide a review of this history, as well as the considerations used to develop race-free equations that are a guidepost for a more equity-oriented, scientifically rigorous future for kidney function estimation and other clinical algorithms and processes in which race may be embedded as a variable.

Comparison of 2021 CKD-EPI Equations for Estimating Racial Differences in Preemptive Waitlisting for Kidney Transplantation

BACKGROUND AND OBJECTIVES

Wait time for kidney transplantation can accrue when GFR is ≤20 ml/min. We examined whether using the race-free 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations to guide preemptive waitlisting could attenuate racial differences in accruable preemptive wait time.

DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS

Our retrospective cohort study included Black or White Chronic Renal Insufficiency Cohort (CRIC) participants who were theoretically eligible for waitlist registration. We used Weibull accelerated failure time models to determine the association between race (Black or White) and time to kidney failure from the qualifying visit when the eGFR by creatinine or creatinine-cystatin C 2021 CKD-EPI equations fell to ≤20 ml/min per 1.73 m2. We then tested for differences in the time ratios from models using the 2021 creatinine- or creatinine-cystatin C-based CKD-EPI equation through a bootstrapping approach.

RESULTS

By the creatinine equation, 472 CRIC participants were theoretically eligible for waitlist registration, and potential preemptive wait time was similar for Black versus White participants (time ratio, 1.05; 95% confidence interval, 0.81 to 1.35). The median wait time by the creatinine equation that could be accrued for Black participants was 23 versus 22 months in White participants. By the creatinine-cystatin C equation, 441 CRIC participants were eligible, and potential wait time was 20% shorter (95% confidence interval, 0.62 to 1.02) for Black than White participants. The median wait time that could be accrued for Black participants was 21 versus 26 months for White participants when using the creatinine-cystatin C equation. Using bootstrapping, the ratio of the time ratio of the models using the creatinine versus creatinine-cystatin C equation was statistically significantly different (ratio of the time ratios = 1.31 with 95% confidence interval, 1.06 to 1.62).

CONCLUSIONS

Use of the 2021 creatinine-based CKD-EPI equation to determine preemptive waitlist eligibility reduced racial differences in preemptive wait time accrual more than use of the creatinine-cystatin C 2021 CKD-EPI equation within a theoretical context.

Short Report: Race and Ethnicity Misclassification in Kidney Transplantation Research

Recently, the misuse of race as a biological variable, rather than a social construct, in biomedical research has received national attention for its contributions to medical bias. In national transplant registry data, bias may arise from measurement imprecision because of the collection of provider-perceived race rather than patients’ own self-report.

Comparing the Net Benefits of Adult Deceased Donor Kidney Transplantation for a Patient on the Preemptive Waiting List vs a Patient Receiving Dialysis

IMPORTANCE

Preemptive kidney transplantation is the preferred treatment for end-stage kidney disease. However, deceased donor (DD) kidneys are limited, and the net benefit of allocating kidneys to a preemptively waitlisted patient rather than to a patient receiving dialysis is unclear.

OBJECTIVE

To estimate the net benefit and costs of allocating kidneys to preemptively waitlisted patients vs those receiving dialysis.

DESIGN, SETTING, AND PARTICIPANTS

This medical decision analytical model used data from the 2020 US Renal Data System to calculate patient survival among waitlisted patients who received a DD kidney transplant. Four patients were simulated, with similar characteristics: (1) a patient on the preemptive waiting list receiving a DD transplant, (2) a patient on the preemptive waiting list never receiving a transplant, (3) a waitlisted patient already receiving dialysis (dialysis vintage <1 year) receiving a transplant, and (4) a waitlisted patient already receiving dialysis (dialysis vintage <1 year) never receiving a transplant. Annual probability of initiating dialysis (for patients 1 and 2) and duration of dialysis (for patients 3 and 4) were varied in sensitivity analyses.

EXPOSURES

Allocating a DD kidney to a patient on the preemptive waiting list vs the same kidney to a patient receiving dialysis for less than 1 year, with similar recipient characteristics.

MAIN OUTCOMES AND MEASURES

Differences in projected quality-adjusted life-years (QALYs) and total costs.

RESULTS

In a simulated patient with a mean start age of 50 years (range, 30-64 years), the patient receiving a preemptive DD transplantation experienced 10.58 (95% CI, 10.36-10.80) QALYs, and the patient on the preemptive waiting list never transplanted experienced 6.83 (95% CI, 6.67-6.99) QALYs. The patient receiving DD transplantation after less than 1 year of dialysis experienced 10.33 (95% CI, 10.21-10.55) QALYs, and the patient receiving dialysis who remained on the waiting list experienced 6.20 (95% CI, 6.04-6.36) QALYs; allocating a DD kidney to the preemptive patient added 3.75 (95% CI, 3.57-3.93) QALYs, whereas allocating the kidney to the patient already receiving dialysis added 4.13 (95% CI, 3.92-4.31) QALYs. While the estimated posttransplant survival was longest for the preemptive transplant recipient, preferentially allocating the kidney to the preemptive patient results in 0.39 (95% CI, 0.49-0.29) fewer QALYs. The net cost of preemptive transplantation was $54 100 (95% CI, $44 100-$64 100) more than transplantation to a waitlisted patient. If the rate of transitioning to dialysis was 20 (rather than 33) events per 100 patient waiting list-years, the net QALYs were -0.67 (95% CI, -0.78 to -0.56). If the patient was receiving dialysis for 3 to 4 years (vs <1 year) the net benefit was not significantly different; however, net costs were considerably higher for the preemptive option.

CONCLUSIONS AND RELEVANCE

In this decision analytic model study, although allocating DD kidneys to patients preemptively was the best option from a patient perspective, allocating DD kidneys to patients receiving dialysis was a better use of a scare resource from a societal perspective.